The two ways it dies
| Failure | Mechanism | Fix |
|---|---|---|
| Free-riding | Contributors notice they're subsidising readers and stop | Contribute-to-access, enforced from week one |
| Unverified errors | A wrong claim spreads with the document's authority | Attribution on everything, plus a source reference |
| Inconsistent quality | Some notes are transcripts, some are three bullets | An agreed minimum format |
| Nobody owns it | The folder rots after week four | One named person per module, rotating |
| It replaces attendance | People stop going because notes exist | Share processed notes, not raw transcripts |
Rules that keep it alive
- 1
Contribute before you can access
The single rule that matters. Everyone submits their week's notes before the folder opens. Without it you get four writers and a readership.
- 2
Attribute everything
Name and date on each document. Not for credit — so that a claim that looks wrong can be queried rather than silently absorbed.
- 3
Cite the slide or page for anything specific
Numbers, definitions, dates. This is what makes an error findable rather than permanent.
- 4
Agree a minimum format in week one
Headings matching lecture sections, key claims marked, questions at the end. Consistency is what makes a pool usable rather than a folder of files.
- 5
Rotate a coordinator per module
One person responsible for chasing and organising, changing every few weeks. Unowned shared resources decay.
Share questions, not just notes
The highest-value thing a cohort can build together isn't a note pool — it's a question bank. Five people writing fifteen questions per lecture produces something nobody could build alone, from the same lectures, weighted by the same lecturer's emphasis.
It also fixes the free-riding problem better than notes do, because writing questions is visibly effortful and the output is obviously useful. How to build a question bank covers the construction; the sharing rules above are what keep it staffed.
What's worth sharing and what isn't
- Worth sharing: processed notes, question banks, past-paper answers with reasoning, reading summaries, and "what the lecturer said would come up".
- Worth sharing carefully: essay plans and structures, which help thinking. Full essays don't, and create real academic-conduct risk for both parties.
- Not worth sharing: raw transcripts, which nobody reads, and unedited generated summaries, which are unverified and spread errors efficiently.
- Never share: anything for an assessed individual submission, or material your department has explicitly restricted. Check the policy rather than assuming — see is using AI for homework cheating for how these rules are usually drawn.
Keep the group small
Four to six people is the range that works. Below four there aren't enough contributors to cover the timetable; above eight the social pressure that makes contribute-to-access work stops functioning, because nobody notices an individual not contributing.
Large cohort-wide pools reliably degrade into a dumping ground with a handful of maintainers. If you want one anyway, run your small group inside it — same study group dynamics as everywhere else.
Verify before you learn from it
Notes written by a classmate carry no more authority than notes written by you, and the shared-document format lends them an unearned air of officialness. That's the mechanism by which one person's misunderstanding becomes the cohort's.
So before anything from a shared pool goes into your cards, check it against the slides — particularly numbers, definitions and anything you find surprising. Surprising claims are either the most valuable thing in the document or an error, and the check takes seconds.
Read shared notes actively
The subtle cost of a good note pool is that reading someone else's clear summary feels like learning and mostly isn't. You get the comprehension without the selection, compression and reformulation that made them useful to write.
The correction is small: after reading a shared note, close it and write three questions from memory. That converts consumption into retrieval and keeps the pool from becoming a very efficient way of feeling prepared — which is the same rereading versus retrieval trap in a social form.